Glenn King of the University of Connecticut Health Center in Farmington and his colleagues recently identified a unique family of toxins in the venom of a funnel-web spider. These neurotoxins are lethal when injected into insect tissues, yet have no effect if eaten by insects or other animals (Nature Structural Biology, vol 7, p 505).
King's team is now engineering the gene for one of these toxins into baculoviruses, common viruses that infect certain moths and butterflies, and have long been used as "biopesticides". When the modified baculovirus infects an insect, the insect's cells should start to produce the toxin, killing it faster than wild viruses. Because the host dies quickly, before much virus can replicate, the modified virus shouldn't persist in the environment, say the researchers.
"I welcome a potentially environmentally friendly pest control but it's abundantly clear we need to be more firm about risk issues," comments George McGavin, an entomologist at Oxford University. "If we are not 100 per cent sure, it shouldn't be in the field."
There have already been several field trials worldwide of baculoviruses given a gene for a scorpion toxin (New Scientist, 21 January 1995, p 6). However, most of the scorpion toxin made in infected insects fails to fold into the correct shape, says King. By contrast, tests in bacteria suggest that almost 100 per cent of the spider toxin should fold properly, making the virus deadlier.
King thinks engineering toxin genes into viruses is preferable to adding them to plants, such as Bt maize. Not only does it mean that people do not have to eat plants that produce insecticidal toxins, but only target insects will be affected, he says. "These viruses can be exquisitely specific, right down to infecting individual species," King claims. "This means that only the pest insects will be killed whilst beneficial insects such as bees remain unaffected."
However, critics fear that the virus will spread into the environment and affect other kinds of butterflies and moths. "A containment environment could not possibly hold a virus," says McGavin, who opposed trials of a scorpion toxin virus in Oxfordshire in the 1990s. "If you could get a specific baculovirus it would be great, but baculoviruses do pass on [to other species]."
There are also fears that the toxin gene might be transferred to other viruses. "There is no instance of a toxin gene jumping from virus A to virus B," says Bruce Hammock of the University of California, Davis, who is also working on modified baculoviruses. "But if it jumped, the new virus would become less effective."
Jenny Cory of the Centre for Ecology and Hydrology in Oxford agrees transfer of the toxin gene is unlikely, but thinks further tests would be helpful. "It's a vicious circle," she says, "you have to do a risk assessment before you do the experiment but we don't know all the risks without doing field experiments in the first place."
New Scientist authors: Mark Robins and Michael Le Page
New Scientist issue: 17th June 2000
PLEASE MENTION NEW SCIENTIST AS THE SOURCE OF THIS STORY AND, IF PUBLISHING ONLINE, PLEASE CARRY A HYPERLINK TO: http://www.newscientist.com